Male Sprague–Dawley rats (220–250 g) were used for all experiments to avoid any possible impact of the estrous cycle on the outcomes of the present study. Animals were obtained from The University of Queensland Biological Resources (UQBR) and were housed in groups of 3–4 per cage in individually ventilated cages in a temperature-controlled facility (22–23°C) with a 12 h/12 h light/dark cycle. Animals were given access to rodent chow and water ad libitum except during behavioral observations. All experimental procedures were approved by the ‘Animal Ethics Committee, The University of Queensland’ and performed with adherence to the ‘Australian Code of Practice for the Care and Use of Animals for Scientific Purposes’ (National Health and Medical Research Council [NHMRC], 2013). After arrival in our facility, rats were acclimatized for approximately one week prior to experimentation. Rats (n = 35) were randomly assigned to one of the following groups: sham, CCI-vehicle or CCI-EMA300 groups (n = 8–15 rats per group).

In rats deeply anaesthetized with 3% isoflurane delivered in oxygen, a unilateral CCI was induced by tying four loose ligatures (4–0 silk sutures; Dysilk Black Braided Siliconized Silk, Dynek, Port Adelaide, SA, Australia) at least ∼1 mm apart around one sciatic nerve (Bennett and Xie, 1988; Bennett et al., 2001). In sham-animals, the sciatic nerve was exposed but not ligated. Each rat received subcutaneous benzylpenicillin (60 mg) as antibiotic prophylaxis. The day of surgery was regarded as day 0.

Calibrated Semmes–Weinstein von Frey filaments (Stoelting Co., Wood Vale, IL, United States) were used to determine paw withdrawal thresholds (PWTs) in the ipsilateral (operated side) and contralateral (non-operated side) hindpaws of all rats. Animals were acclimatized for at least 20 min in wire mesh metabolic cages (20 cm × 20 cm × 17 cm) prior to assessment of bilateral PWTs. The von Frey filaments were applied to the plantar surface of the hindpaws to deliver forces in the range 2–20 g at intervals of 2 g using the up-down protocol (Chaplan et al., 1994; Smith et al., 2013a). Animals were randomized prior to behavioral testing by an independent person to ensure that behavioral assessments were conducted in a blinded manner. Temporal development of mechanical allodynia in the ipsilateral hindpaw was assessed once weekly over the 14–17 day study period. The PWTs of the contralateral hindpaws were also assessed during each testing session.

Between days 14–17 post-surgery, CCI-rats with fully developed ipsilateral hindpaw hypersensitivity (PWTs ≤ 6g) were randomized to receive a single intraperitoneal (i.p.) bolus dose of the sodium salt of EMA300 (also known as PD121981) at 10 mg kg^-1 (n = 15) or vehicle (water for injection; n = 8). The bilateral hindpaw PWTs were assessed pre-dose and at ∼1 h post-dose, the time of the EMA300 peak-effect (Smith et al., 2013a,b), following administration of a bolus dose of either EMA300 or vehicle. Sham (control) rats received no treatment, and their PWTs were assessed at similar times to that of the CCI-rats for comparison. All dosing and behavioral assessments were performed by independent personnel in a blinded manner. Specifically, administration of the test-item (EMA300) or vehicle to rats was done by one person whereas the behavioral testing was done by another independent person.

Following assessment of the bilateral hindpaw PWTs at ∼1 h post-EMA300 or vehicle administration, rats were euthanized with an overdose of pentobarbitone (Lethabarb^®, Virbac, Milperra, NSW, Australia) that was administered i.p. at a dose of ∼300 mg kg^-1 (i.e., 1 mL kg^-1 of the stock solution prepared as 325 mg mL^-1) per rat (Smith et al., 2013b). A group of age-matched sham-control rats (n = 12) was also euthanized in a similar manner. Following laminectomy, the ipsilateral lumbar (L4–L6) DRGs were removed from all rats. These tissues were then processed for ex vivo investigations using immunohistochemistry (IHC) as well as western blotting (WB). For IHC, rats were perfused with 4% paraformaldehyde (Sigma–Aldrich, Sydney, NSW, Australia) in ice-cold 1x PBS (pH 7.4; ∼150–200 mL per rat) prior to extraction of the lumbar DRGs. Serial cross-sections (8–10 μm thick) were obtained from DRGs embedded in Optimal Cutting Temperature (O.C.T) compound (ProSciTech, QLD, Australia). Superfrost Plus^® slides (Thermo Fisher Scientific, Scoresby, VIC, Australia) were used to collect the cryosections. The slides were stored at -20°C until further processing for IHC. For WB, the lumbar DRG tissues were collected without perfusion and were snap-frozen using a dry-ice/ethanol slurry followed by immediate transfer to a -80°C freezer until further experimentation.

The lumbar DRG sections from all treatment groups were air-dried prior to immunostaining followed by washing with PBS containing 0.05% Tween 20 (Sigma-Aldrich, Sydney, NSW, Australia) (referred herein as PBST) for 2 × 10 min. Subsequently, the sections were incubated for 1–2 h in blocking buffer containing 10% normal goat serum (NGS) (Invitrogen, Mulgrave, VIC, Australia) or 5% Bovine Serum Albumin (BSA) (Sigma–Aldrich) as appropriate for each IHC antibody (Muralidharan et al., 2014; Khan et al., 2015). Later, the sections were incubated with appropriate primary antibodies for at least 15–20 h at ∼4–8°C. Subsequently, the sections were incubated with the corresponding secondary antibodies for ∼1 h at room temperature. The sections were then stained with 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI, Invitrogen) to stain neuronal nuclei followed by mounting with ProLong^® Gold anti-fade reagent (Invitrogen). Sections were washed with PBST for at least 2 × 10 min after all antibody incubation steps. The control sections were stained with omission of primary antibodies. To identify the cellular co-localization of Ang II and the AT₂ receptor in ipsilateral lumbar DRG sections, we used simultaneous immunostaining of these markers with NeuN, CD3, and glial fibrillary acidic protein (GFAP) that are specific markers for neurons, T-cells and SGCs, respectively. The antibodies against the AT₂ receptor and GFAP were used separately on adjacent sections to avoid cross-reactivity as both antibodies were raised in the rabbit. The images were snapped in the exact location in the region of interest and were overlaid to observe the co-localization. Additionally, we performed co-localization immunostaining for Ang II with the AT₂ receptor and the AT₂ receptor with NGF. Prior to the main study, pilot experiments were conducted to optimize the dilutions of each antibody and determine the optimal incubation time, blocking buffer, washing steps and DAPI staining.

The primary antibodies used include, polyclonal guinea pig anti-Angiotensin II (1:200, Peninsula laboratory, CA, United States); polyclonal rabbit anti-AT₂ receptor (1:500, ab19134, Abcam, Cambridge, MA, United States), monoclonal rabbit anti-AT₂ receptor (1:500, ab92445, Abcam), monoclonal rat anti-CD3 (1:100, Serotec, Kidlington, United Kingdom); monoclonal rabbit anti-CD3 (1:100, Abcam); monoclonal mouse or rabbit anti-pp38 MAPK and anti-pp44/pp42 MAPK (Cell Signaling Technology, Inc., Danvers, MA, United States); monoclonal rabbit anti-GFAP (1:500, Cell Signaling); rabbit polyclonal anti-NGF (1:500, H-20, Santa Cruz Biotechnology, Santa Cruz, CA, United States); polyclonal sheep anti-NGF (1:500, Abcam); monoclonal rat anti-TrkA (1:50, R&D Systems, Minneapolis, MN, United States) and mouse monoclonal anti-NeuN Alexa Fluor^®488 conjugated, clone A60 (1:100, Millipore, Temecula, CA, United States). For most of the IHC markers, at least two different antibodies were used as described above to validate the specificity of the immunostaining. Secondary antibodies used include, Alexa Flour 633 goat anti-guinea pig (1:500, Invitrogen); Alexa Flour 546 goat anti-guinea pig (1:800, Invitrogen); Alexa Flour 555 goat anti-sheep (1:800, Invitrogen); Cy3-goat anti-rat (1:800, Jackson ImmunoResearch, Westgrove, PA, United States); Cy3-goat anti-rabbit (1:800, Jackson); Alexa Flour 488 goat anti-rat (1:500, Invitrogen) or Alexa Flour 488 goat anti-rabbit (1:500, Invitrogen). All antibodies were diluted in PBST containing 2% NGS or BSA blocking buffer as appropriate. An Axioskop 40 microscope (Carl Zeiss, Göttingen, Germany) attached to an Axiocam MRm camera (Carl Zeiss) was used for imaging sections at a fixed exposure time that was optimized for each stain in preliminary experiments (Khan et al., 2015).

The ipsilateral lumbar DRGs (L4–L6) from each rat in the respective experimental cohort were pooled and homogenized on ice for protein extraction using radio-immunoprecipitation (RIPA) buffer. The buffer was prepared by dissolving 10 mM Tris; 100 mM NaCl, 1%; 1 mM ethylenediaminetetraacetic acid (EDTA); 1mM ethylene glycol tetraacetic acid (EGTA); 1 mM sodium fluoride (NaF); 20 mM tetrasodium pyrophosphate (Na₄P₂O₇); sodium orthovanadate (Na₃VO₄); 0.1% sodium dodecyl sulphate (SDS), and 0.5% sodium deoxycholate in MilliQ water followed by addition of 1% Triton X-100 and 10% Glycerol; pH 7.4). The buffer was also supplemented with 1% of each Halt^TM Phosphatase Inhibitor Cocktail (Thermo Fisher Scientific, Scoresby, VIC, Australia) and protease Inhibitor Cocktail (Sigma–Aldrich). The protein concentrations were determined with the Bradford method (protein assay kit, Bio-Rad Laboratories, Milan, Italy). From each extract, 30 μg of total protein was loaded onto the 4–20% Mini-PROTEAN^® TGX Stain-Free^TM precast polyacrylamide gels (Bio-rad, Gladesville, NSW, Australia). Each gel was then transferred to a 0.45 μm Immobilon-FL PVDF membrane (Millipore) and incubated in Odyssey^TM blocking buffer (Li-cor, Lincoln, NE, United States) for at least 1 h. For phospho-specific antibodies, the blocking buffer comprised 5% BSA/0.1% Tween-20. The membranes were incubated overnight at ∼4°C with specific antibodies against the AT₂ receptor (1:500, ab92445, Abcam), GFAP (1:300, Cell signaling), NGF (1:500, H-20), TrkA (1:500, R&D Systems), phospho-p38 MAPK (pp38 MAPK), phospho-p44/pp42 MAPK, p38 MAPK (total p38 MAPK), and p44/p42 MAPK (total extracellular signal-related kinases (ERK)) (each 1:1000). The sources of primary antibodies are the same as listed in section “Immunohistochemical Staining”. Monoclonal mouse anti-beta actin (1:10,000, Abcam) (1:8000, Novus, Littleton, CO, United States) was used as the loading control. Following washing with PBST (3 × 5 min), the membranes were incubated in the appropriate IRDye^® secondary antibodies (Li-cor) (1:8000) for ∼1 h at room temperature. The Odyssey^® CLx imaging system (Li-cor) was used to visualize the membranes.

Human embryonic kidney (HEK) cells obtained from the American Type Culture Collection (ATCC, VA United States) were transfected with rat AT₁ receptor complementary DNA (cDNA) (HEK-AT₁) alone or rat AT₂ receptor cDNA (HEK-AT₂) alone using Lipofectamine^® 2000 as per the manufacturer’s protocol. Dulbecco’s Modified Eagle’s Medium (DMEM) (Invitrogen) supplemented with 10% Fetal bovine serum (FBS) (Invitrogen) was used for culturing these cell lines. For transfected cell-lines, G418 (Geneticin) (500 mg mL^-1) was added to the growth medium as a selective agent for generating cell lines stably expressing either the AT₂ receptor or the AT₁ receptor. These stable cell lines were maintained at 37°C in a 5% humidified CO₂ incubator. To assess if the disparity between the molecular mass of the AT₂ receptor (kDa) from the DRG tissues and that from HEK-AT₂ cells were potentially related to different degrees of N-glycosylation, the cells were incubated with tunicamycin (TNM) (Sigma–Aldrich), an inhibitor of N-glycosylation. Specifically, the HEK-AT₂ cells were incubated with TNM at 1 or 3 μg mL^-1 for 48 h, conditions previously shown by others to not affect cell viability (Servant et al., 1996). The HEK-AT₂- as well as HEK-AT₁ -cells not treated with TNM served as positive and negative control groups, respectively. The cells were lysed with RIPA buffer and processed for WB analysis using the monoclonal rabbit anti-AT₂ receptor antibody (ab92445, Abcam) as described in preceding section.

In order to confirm the specificity of the monoclonal rabbit anti-AT₂ receptor (Abcam; ab92445), HEK-cells stably expressing either the AT₂ or the AT₁ receptor were cultured using the identical procedure described in preceding section. Briefly, the cells were grown to 50–70% confluency on sterilized coverslips placed in 6-well plates. Following harvest, the cells were fixed with ice-cold 100% methanol for 5 min at room temperature. Subsequently, the cells were washed with PBST (3 × 5 min) and blocked with 10% NGS in PBST for 30 minutes at room temperature. The cells were then incubated overnight with each monoclonal rabbit anti-AT₂ receptor antibody (ab92445 and ab19134, Abcam). For visualization, Alexa Flour 546 goat anti-rabbit (1:800, Invitrogen) was used and it was incubated with cells for at least 1 h. The cells were then incubated with DAPI and were mounted using ProLong^® Gold anti-fade reagent prior to imaging using similar washing steps as described in section “Immunohistochemical Staining”. The cells incubated with only secondary antibody served as an internal experimental control. Un-transfected HEK-cells were used as an additional negative control. The HEK-AT₂ cells stained with anti-AT₂ receptor antibody (Abcam, ab19134), the specificity of which was previously validated using adrenal and DRG tissues from AT₂ receptor knockout mice (Smith et al., 2013b), served as a positive control.

Behavioral data (PWT values) are presented as mean (±SEM) of three measurements taken at least 5 min apart individually for the ipsilateral and contralateral hindpaws. For IHC, the Axiovision Rel. v4.8 software (Carl Zeiss, Göttingen, Germany) was used to determine the mean fluorescence intensity. The IHC quantification was performed on at least four lumbar L4–L6 DRG sections (>100 μm apart) obtained from 3 to 4 rats in each treatment group. IHC quantification data is expressed as fold-changes in fluorescence intensity relative to that for the corresponding sections from sham-rats. ImageJ 1.48v (NIH, United States) was used for counting CD3⁺ T-cells with or without Ang II co-expression. For western blot analysis, Image Studio Lite v4.0 software (Li-cor, Lincoln, NE, United States) was used to quantify the band intensities for each treatment group using 3–4 rats per group. Band intensities were normalized relative to beta-actin and the results are shown relative to the corresponding data from sham-control rats.

To analyze the pain behavioral end-points, we used two-way analysis of variance (ANOVA) with a post hoc Bonferroni test for multiple comparisons. The one-way ANOVA followed by Tukey’s multiple comparison test was used to analyze the between-group differences for IHC and WB. GraphPad Prism^TM v7 (GraphPad Software, La Jolla, United States) was used for statistical analyses. The statistical significance criterion was P ≤ 0.05.

Mechanical allodynia (PWTs ≤ 6 g) in the ipsilateral hindpaws of CCI-rats was fully developed by day 14–17 post-surgery (Figure 1A). By contrast, mechanical allodynia did not develop in the contralateral hindpaws of CCI-rats or in the bilateral hindpaws of the sham-group rats (Figures 1A,B). Following administration of a single i.p. bolus dose of EMA300 (10 mg kg^-1) to CCI-rats with fully developed ipsilateral hindpaw hypersensitivity, there was significant [F_(2,3,6/96) = 51.66, 115.7, 43.65; P ≤ 0.05] relief of mechanical allodynia in the ipsilateral hindpaws at ∼1 h post-dosing (time of peak-effect), whereas vehicle (saline)-treatment was inactive (Figure 1A). These findings are in agreement with previous work from our laboratory (Smith et al., 2013a,b). Importantly, single i.p. bolus doses of EMA300 (10 mg kg^-1) did not result in any discernible central nervous system (CNS) side-effects, for example, sedation, or neuro-excitatory behaviors, during all assessments.

Using IHC, our findings show that Ang II and the AT₂ receptor are co-localized with CD3⁺ T-cells and SGCs (GFAP) in the ipsilateral lumbar DRG sections from CCI-rats exhibiting neuropathic pain behavior (Figures 2, 3). Additionally, in agreement with our previous observations (Smith et al., 2013b; Muralidharan et al., 2014), Ang II and/or the AT₂ receptor were also expressed by subsets of sensory neurons marked by neuronal nuclei protein (NeuN) in addition to NGF in the ipsilateral lumbar DRGs of CCI-rats (Figures 2, 3).

Between days 14–17 post-surgery, there was a significant increase in the number of Ang II-expressing CD3⁺ T-cells (∼9.8-fold; F_(2,33) = 64.1; P ≤ 0.05) as well as the total number of CD3⁺ T-cells (∼5.3-fold; F_(2,33) = 53.37; P ≤ 0.05) in the ipsilateral lumbar DRG sections of vehicle-treated CCI-rats relative to sham-group rats (Figure 4). Interestingly, administration of a single i.p. bolus dose of EMA300 (10 mg kg^-1) to CCI-rats significantly reduced the otherwise increased number of CD3⁺ T-cells expressing Ang II (F_(2,33) = 64.1; P > 0.05) and the total number of CD3⁺ T-cells (F_(2,33) = 53.37; P ≤ 0.05) at its time of peak analgesic effect (∼1 h post-dosing). In particular, the number of Ang II-expressing CD3⁺ T-cells in the ipsilateral lumbar DRGs of CCI-rats treated with EMA300 (10 mg kg^-1) was reduced to match the number in the corresponding sections from the sham-group of rats (Figure 4).

Herein, we found that the polyclonal rabbit anti-AT₂ receptor antibody (ab19134), previously validated for IHC using adrenal gland and DRG sections from AT₂ receptor knockout mice (Smith et al., 2013b), is not compatible for use with western blotting (data not shown). Therefore, we used a monoclonal rabbit anti-AT₂ receptor antibody (ab92445) for western blots and validated its specificity for the AT₂ receptor using western blotting and immunocytochemistry (ICC) (Supplementary Figure S1). Our results demonstrate that the monoclonal rabbit anti-AT₂ receptor antibody (ab92445) bound specifically to HEK-cells stably expressing the AT₂ receptor, but not to un-transfected HEK-cells or to HEK-cells stably expressing the AT₁ receptor (Supplementary Figure S1). In ICC, the polyclonal rabbit anti-AT₂ receptor antibody (ab19134) was used as a positive control antibody to corroborate the results obtained using the monoclonal rabbit anti-AT₂ receptor antibody (ab92445). During the western blot analysis of the HEK-AT₂ cells, the expression of the AT₂ receptor was observed at ∼90 kDa (Supplementary Figure S1). In contrast, the AT₂ receptor was expressed at ∼50 kDa for the ipsilateral lumbar DRGs samples from CCI- and sham-rats (Figure 5C). A possible explanation of this paradoxical observation may be due to differential levels of glycosylation, which are frequently associated with the AT₂ receptor (Servant et al., 1994, 1996). In support of this notion, following de-glycosylation with tunicamycin (TNM) at 1 and 3 μg mL^-1 for 48 h, the band for the AT₂ receptor shifted from 90 to ∼50 kDa, to match that of the AT₂ receptor in DRG tissue samples (Supplementary Figure S1).

In agreement with previous work from our laboratory (Smith et al., 2013b), there was a ∼4.7-fold increase in the Ang II IF (F_(2,33) = 58.65; P ≤ 0.05) in the ipsilateral lumbar DRGs of CCI-rats with fully developed ipsilateral hindpaw hypersensitivity c.f. the corresponding sections from sham-group rats (Figures 5A,B). At the time of peak analgesia evoked by a single i.p. bolus dose of EMA300 at 10 mg kg^-1, the elevated expression levels of Ang II IF in the CCI-rats were markedly decreased (F_(2,33) = 58.65; P ≤ 0.05) to match the corresponding levels in sham-rats (Figures 5A,B).

There was no significant difference (P > 0.05) in the expression levels of the AT₂ receptor in sections of ipsilateral lumbar DRGs from CCI-rats administered a single i.p. bolus dose of either EMA300 or vehicle using IHC (F_(2,44) = 0.6060, Figures 5A,B) and western blot (F_(2,6) = 0.6946, Figures 5C,D).

Herein, total NGF IF did not differ significantly (F_(2,33) = 3.695; P > 0.05) in the ipsilateral lumbar DRG sections from CCI- and sham-rats (Figures 6A,B). However, there were significant differences for the individual NGF isoforms when assessed by WB such that these isoforms were significantly reduced (pro-NGF, F_(2,9) = 13.12; mature-NGF, F_(2,9) = 5.944; P ≤ 0.05) in vehicle-treated CCI-rats compared with sham-rats (Figures 6C,D). Notably, we show for the first time that a single i.p. bolus dose of EMA300 completely restored the otherwise significantly reduced (F_(2,9) = 11.95; P ≤ 0.05) mature-NGF expression levels in the ipsilateral lumbar DRGs of CCI rats at the time point of its peak analgesia (Figures 6C,D). Additionally, there was no significant change in pro-NGF expression in the ipsilateral lumbar DRGs of CCI-rats treated with EMA300 when compared with the vehicle treated CCI group (Figures 6C,D).

In the somatosensory DRG neurons from animal models, the high affinity NGF receptor, TrkA, is implicated in the pathophysiology of neuropathic pain (Grelik et al., 2005; Ugolini et al., 2007). In agreement with previous reports by others (Fang et al., 2005), we also observed that there was an ∼3-fold increase (F_(2,28) = 14.37; P ≤ 0.05) in TrkA expression in the ipsilateral lumbar DRG sections of CCI-rats with fully developed hindpaw hypersensitivity c.f. the corresponding sections from sham-rats (Figures 6A,B). However, the elevated TrkA expression levels were not altered by a single bolus dose of EMA300 (Figures 6A,B). These findings suggest that the acute anti-allodynic effects of EMA300 in CCI rats is not transduced by a change in TrkA expression.

Consistent with previous work by others (Hu et al., 2007; Liu et al., 2016), we found a significant increase in GFAP IF (IHC ∼3.1 fold, F_(2,33) = 15.23; WB ∼1.7-fold, F_(2,9) = 5.398; P ≤ 0.05) in sections of the ipsilateral lumbar DRGs of vehicle-treated CCI rats relative to the corresponding sections from sham-rats (Figure 7). However, at the time of peak effect of a single i.p. bolus dose of EMA300 (10 mg kg^-1), there was insignificant attenuation (P > 0.05) of GFAP IF suggesting that the anti-allodynic effect of acutely administered EMA300 in CCI-rats is not mediated by regulation of GFAP in the ipsilateral lumbar DRGs of CCI-rats (Figure 7).

Consistent with previous work from our laboratory (Smith et al., 2013b), there was a significant increase in P-p38 MAPK (IHC ∼3.7-fold, F_(2,6) = 21.01; WB ∼2.2-fold, F_(2,6) = 21.01; P ≤ 0.05) and P-p44/p42 MAPK (IHC ∼3-fold, F_(2,6) = 33.58; WB ∼4.1-fold, F_(2,6) = 33.58; P ≤ 0.05) in the ipsilateral lumbar DRGs of vehicle-treated CCI rats compared with the corresponding values for sham rats (Figure 8). Notably, the augmented levels of P-p38 MAPK and MAPK were reduced in CCI rats following a single bolus i.p. dose of EMA300 to match the respective level in sham-rats (Figure 8).